Download Final Exam Study Guide

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Foundations of statistics wikipedia , lookup

Psychometrics wikipedia , lookup

Student's t-test wikipedia , lookup

Analysis of variance wikipedia , lookup

Resampling (statistics) wikipedia , lookup

Misuse of statistics wikipedia , lookup

Transcript
You can use the following resources in the exam:
One page of notes (you can use both sides)
A calculator
A table of areas under the standard normal distribution and a table of critical t-scores, and
F ratios (We will provide these for you)
We will give you scrap paper for doing computations
RESEARCH DESIGN
You should understand the following concepts and be able to identify them given a
hypothetical research project and to apply them to the interpretation of an
experiment’s results:
Experiments vs. observational studies (and their pros and cons)
Independent vs. dependent variables
Between subject vs. within subject designs (and their pros and cons)
What it means to counterbalance the order of a within-subject manipulation
You should understand the importance of random, representative sampling in
research and of random assignment to experimental conditions in experiments.
DESCRIPTIVE STATISTICS
Given a hypothetical set of data you should be able to identify which method of
visualization is most appropriate (bar graph, histogram, polygon, or scatter plot),
which measure of central tendency is most appropriate (mean, median, or mode),
and which measure of dispersion is most appropriate (standard deviation,
interquartile range, or entropy). You should also be able to explain why your
chosen type of graph or descriptive statistic is the most appropriate.
You should be able to identify data as:
Nominal
Ordinal
Interval/Ratio
You should be able to label data distributions as:
Normal
Positively Skewed
Negatively Skewed
Bimodal
Supergaussian
Subgaussian
You need to know how to read the following graphs:
Histogram
Bar graph
Frequency polygon
Boxplot
Scatterplot
Note: You will NOT be asked to produce any graphs. But you need to be able to
interpret them, to know when a particular type of graph is an appropriate way to visualize
the data, and to identify misleading graphs.
From a scatter plot, you should be able to determine if linear or rank correlation is
a better way of quantifying the relationship between two variables. You should also
be able to guess the linear correlation coefficient Pearson’s r from a scatter plot.
Given Pearson’s r for a data set, you should be able to interpret it (e.g., if two
variables are correlated with an r of .56, what does that mean?).
You will NOT need to know how to compute the mode, interquartile range, range,
or entropy.
You will NOT need to compute frequency distributions.
PROBABILITY & INFERENTIAL STATISTICS
Given a hypothetical data set and research question, you should be able to identify
which hypothesis test is most appropriate to analyze the data. The possible options I
expect you to know are:
z-test
one sample t-test
binomial/sign test
repeated measures t-test
independent samples t-test
Mann-Whitney U test
t-test of Pearson’s linear correlation coefficient r
rank correlation (e.g., Kendall’s tau or Spearman’s rho)
independent samples one factor ANOVA
repeated measures one factor ANOVA
independent samples two-factor ANOVA
linear regression
“none of these”
You should know how to compute a z-test, one sample t-test, and a repeated
measures t-test (this includes deriving p-values, confidence intervals, and Cohen’s
d). You should be able to fill out hypothesis testing forms for JUST these 3 tests.
You should be able to complete an ANOVA table given all the sums of squares (i.e.,
you will have to figure out the degrees of freedom, mean squares, F ratios, p-values
of F ratios, and determine if an F ratio is significant at a given alpha level—see
Question 11 on the practice final for example).
You should be able to compute the eta squared (η2) and partial eta squared (ηp2) for
ANOVA effects and interpret Tukey’s HSD to find out which pairs of cells differ.
You will NOT have to perform Tukey’s HSD.
You should be able to read two factor ANOVA plots to determine which main
effects appear to be significant and if the interaction appears to be significant (see
HW 9 Problem 9.D on WebCT for an example).
You should be able to read the results of a hypothesis test and to think critically
about them (e.g., What do the results suggest about the likelihood that the null or
alternative hypothesis is true? What do the results suggest about the size of any
potential effect?)
You should understand and be able to explain the following concepts:
Population distribution vs. the sampling distribution of the mean (Chapters 8-9)
The central limit theorem
What it means to reject or retain the null hypothesis
Type I and Type II error and ways to reduce/increase the probability of making them
The alpha level of a test
The p-value of a test statistic
Cohen’s d
The power of a hypothesis test
The effect of sample size on hypothesis tests and estimation
The difference between parametric (e.g., t-tests) and nonparametric (e.g., the MannWhitney U test) tests.
Why multiple comparisons are problematic
The purpose of following up ANOVAs with a multiple comparisons test like Tukey’s
HSD
eta squared (η2) and partial eta squared (ηp2)
The concept of “main effects” and “interactions” in a two factor ANOVA
The purpose of following up two factor ANOVAs with a “simple effects test”
You do NOT need to know how to compute:
an independent samples t-test
a binomial/sign test
Mann-Whitney U test
Wilcoxon T
Power curves
correlation coefficients
linear regression
sums of squares for an ANOVA
simple effects tests
the formula for binomial distributions
Basic probability problems like the first questions 4,5, and 6 on practice midterm II
Conditional Probability
Bayes’s rule
Tukey’s HSD
Again, you do NOT need to know how to fill out hypothesis testing forms (except for
the 3 tests I expect you to be able to compute: z-test, one sample t-test, repeated
measures t-test.